1. Expression of Long-Term Plasticity at Individual Synapses in Hippocampus Is Graded, Bidirectional, and Mainly Presynaptic: Optical Quantal Analysis 
Ryosuke Enoki, Yi-ling Hu, David Hamilton, Alan Fine 
Neuron 
Volume 62, Issue 2, Pages (April 2009)
DOI: /j.neuron
Copyright © 2009 Elsevier Inc. Terms and Conditions

2. Figure 1 LTP and LTD Are Not Binary Processes at Individual Synapses
(A) A CA1 pyramidal neuron, filled with the fluorescent Ca2+ indicator Oregon Green 488 BAPTA-1 via sharp intracellular microelectrode. Schaffer collateral activation via a stimulating electrode (SE) in stratum radiatum (sr) evoked EPSPs recorded via the somatic microelectrode (not visible), as well as calcium transients in an apical dendritic segment (region of interest [ROI] indicated by the white box, seen at higher magnification in [B]).
(B) Evoked postsynaptic calcium transients (EPSCaTs) were restricted to an individual dendritic spine (arrowhead), seen below at higher magnification in video frames at rest and immediately after synaptic activation.
(C) EPSCaTs were monitored with higher temporal resolution via line-scan imaging across the spine and subjacent dendritic shaft (line scan trajectory indicated by vertical white line in the upper image of [B]); time of stimulus application is indicated by dotted vertical line across each line scan image and extending to the unaveraged quantified fluorescence (upper) and simultaneously-recorded intracellular voltage (middle) traces. Successful synaptic transmission (left), visible as a fluorescence increase in the horizontal streak corresponding to the active synapse (arrowhead), can be clearly distinguished from transmission failure (right; EPSP during transmission failure at this synapse is due to successful transmission at some of the other synapses activated by the same extracellular stimulus). Mean fluorescence traces from this spine during successful transmissions (left; n = 20) and transmission failures (right; n = 20) are shown in the lower traces enclosed by dotted lines. Increasing fluorescence intensity in these and other images is represented by colors from black through red to yellow. sp, stratum pyramidale; so, stratum oriens; inset cartoon shows position of this neuron and electrodes within the hippocampal slice.
(D) EPSCaT amplitudes plotted for each stimulus before and 30–45 min after two successive LTP inductions (upward open arrows); dotted line is the limit of the (unstimulated) fluorescence noise distribution and thus the threshold for attribution of transmission success. Incidence of successes (red) provides a direct measure of transmitter release probability (pr), which increased from 0.20 ± 0.08 to 0.42 ± 0.09 after the first LTP induction and then to 0.70 ± 0.10 after the second potentiation. Traces show average EPSPs over each plotted set of trials, increasing from 2.80 ± 0.26 mV baseline to 4.42 ± 0.22 mV after the first LTP and to 5.82 ± 0.31 mV after the second. Values of pr and EPSP for this and five other experiments (black) are shown, with significance of comparisons between weighted means (blue), in (E) and (F), respectively. Error bars in these and all other figures denote standard error of the mean.
(G) EPSCaT amplitudes and mean EPSPs (as in D) from an experiment in which a synapse was subjected to two successive inductions of LTD (downward open arrows), resulting in successive reduction of EPSP from 6.61 ± 0.10 mV baseline to 5.40 ± 0.19 mV after the first induction and to 4.73 ± 0.22 mV after the second, and of pr from 0.64 ± 0.08 to 0.34 ± 0.08 and subsequently to 0.19 ± Values of pr and EPSP for this and 4 other experiments (black) are shown, with significance of comparisons between weighted means (blue), in (H) and (I), respectively.
Neuron  , DOI: ( /j.neuron )
Copyright © 2009 Elsevier Inc. Terms and Conditions

3. Figure 2
Individual Synapses Can Sustain Bidirectional Long-Term Plasticity
(A) EPSCaT amplitudes and mean EPSPs (as in Figure 1) from an experiment in which a synapse was initially potentiated and then depressed: initial LTP induction increased pr from a baseline of 0.20 ± 0.09 to 0.60 ± 0.11, with mean EPSP amplitude increased from 1.30 ± 0.09 to 2.50 ± 0.11 mV; subsequent LTD/depotentiation induction reduced pr to a new value of 0.45 ± 0.11 and the EPSP amplitude to 2.01 ± 0.10 mV. Values of pr and EPSP for this and seven similar “up-down” experiments (black) are shown, with significance of comparisons between weighted means (blue), in (B) and (C), respectively.
(D) Synapses can sustain bidrectional plasticity regardless of the direction of the initial change. Here, a synapse underwent LTD first and then LTP; pr went from 0.97 ± 0.03 baseline to 0.47 ± 0.09 and then to 0.83 ± 0.07, with corresponding changes in EPSP from 1.93 ± 0.05 mV to 1.58 ± 0.04 mV and then to 4.41 ± 0.16 mV. (E) and (F) show values of pr and EPSP for this and four similar “down-up” experiments (black), with significance of comparisons between weighted means (blue).
Neuron  , DOI: ( /j.neuron )
Copyright © 2009 Elsevier Inc. Terms and Conditions

4. Figure 3
Individual Synapses Sustain Multiple Levels of Potentiation and Depression
(A) EPSCaT amplitudes and mean EPSPs (as in Figure 1) from an experiment in which a synapse was sequentially potentiated twice, then depressed, and then potentiated again. These manipulations were associated with sequential changes of pr from a baseline of 0.10 ± 0.06 to 0.38 ± 0.11, 0.57 ± 0.10, 0.05 ± 0.05, and 0.26 ± 0.10, with corresponding changes in mean EPSP amplitude from 3.92 ± 0.20 mV to 8.70 ± 0.33, ± 0.28, 6.18 ± 0.24, and 7.77 ± 0.23 mV.
(B) Sequential values of pr are plotted for this (black) and four other synapses in independent experiments, in response to successive application of LTP-inducing (solid lines) and LTD-inducing (dashed lines) stimuli. All changes in pr are significant, p < 0.05.
Neuron  , DOI: ( /j.neuron )
Copyright © 2009 Elsevier Inc. Terms and Conditions

5. Figure 4
Subtractive Analysis of Unitary EPSPs Indicates LTP at Schaffer Collateral Synapses Is Expressed Mainly via Changes in Transmitter Release Probability
(A) EPSCaT amplitudes (above, as in Figure 1) and EPSP amplitudes (below) recorded before and after LTP induction (open upward arrow). EPSP amplitudes are color-coded depending upon whether they corresponded to EPSCaT successes (red) or failures (black).
(B) Mean EPSPs corresponding to EPSCaT successes (red) and failures (black). Although the EPSPs result from activation of several synapses, the difference between these means reflects the mean unitary EPSP (green) from the imaged active synapse. Traces shown are means before (“Baseline,” left) and 20–60 min after (right) LTP induction. LTP resulted in large increases in the overall mean EPSP (from 3.96 ± 0.09 to 6.95 ± 0.08 mV) and pr at the imaged synapse (from 0.22 ± 0.04 to 0.86 ± 0.03). The unitary EPSP due to this synapse, however, was only slightly changed (from 0.76 ± 0.20 to 0.85 ± 0.22 mV).
(C) Values of (from left to right) overall EPSP, pr, EPSP grouped according to transmission success (S) or failure (F), and unitary EPSP from the imaged synapse, for this and three other experiments (black) are shown before and after LTP, with significance of comparisons between weighted means (blue). LTP induction led to significant and corresponding increases in the (multisynaptic) EPSP and pr, with no significant change in the unitary EPSP.
Neuron  , DOI: ( /j.neuron )
Copyright © 2009 Elsevier Inc. Terms and Conditions

6. Figure 5
Direct Recording of Unitary EPSPs under Optically Confirmed Minimal Stimulation Also Shows that LTP at Schaffer Collateral Synapses Is Expressed Mainly via Changes in Transmitter Release Probability
(A) Representative successive traces showing the perfect correspondence between success or failure of EPSCaTs (left) and EPSPs (right) in an experiment both before (Baseline) and after LTP. This constant correspondence provides strong evidence that the stimulus in this experiment activated only the imaged synapse (i.e., that the somatically-recorded EPSP is the unitary EPSP due to the imaged synapse), and that EPSCaTs are reliable reporters of presynaptic glutamate release.
(B) EPSCaT and EPSP amplitudes, as in Figure 5, recorded from this synapse before and after LTP induction (open upward arrow). LTP induction increased pr (from 0.55 ± 0.07 to 0.80 ± 0.06) but not the unitary EPSP (from 0.99 ± 0.05 to 0.81 ± 0.05 mV).
(C) Values of pr and unitary EPSP from the imaged synapse for this and 2 other experiments (black) are shown before and after LTP, with significance of comparisons between weighted means (blue). LTP induction led to significant increases in pr, with no significant change in unitary EPSP.
Neuron  , DOI: ( /j.neuron )
Copyright © 2009 Elsevier Inc. Terms and Conditions

7. Figure 6 EPSCaT-Generating Synapses Are Major Contributors to the EPSP
(A) Exhaustive search of the dendritic arbor of the illustrated neuron during afferent stimulation (average EPSP 5.28 ± 0.20 mV) revealed 8 synapses generating EPSCaTs (abbreviations as in Figure 1). Their sites (a-g) are shown at higher magnification in (B), along with representative line scans demonstrating EPSCaTs in the activated spines, and the observed values of pr at each synapse. The weighted mean pr for these synapses was 0.43 ± 0.06.
(C) Frequency histogram showing values of pr at the 38 EPSCaT-generating synapses observed by exhaustive search of this and 4 other CA1 pyramidal neurons in separate experiments; overall mean pr = 0.40 ± 0.02.
(D) Plot of the number of EPSCaT-generating synapses in each of these experiments; a mean of 7.60 ± 0.93 such synapses were observed on each cell, in response to Schaffer collateral stimulation that evoked an average EPSP of 4.24 ± 0.06 mV.
(E) Plot of estimated mean contribution of EPSCaT-generating synapses to the EPSP in each of these experiments, assuming linear summation of unitary EPSPs with mean amplitudes for these synapses equal to that of the overall population we studied (0.63 ± 0.07 mV).
Neuron  , DOI: ( /j.neuron )
Copyright © 2009 Elsevier Inc. Terms and Conditions